Best MCPs for Cursor: Supercharge Your AI Coding Environment
Cursor is already a powerful AI coding tool. MCPs make it dramatically better by adding real tool access — accurate docs, live database queries, GitHub integration, and web search. Here are the ones worth installing.
Why Cursor + MCPs?
Cursor has excellent built-in context features — codebase indexing, file references, and rules. But these only cover what is already in your project. MCPs extend Cursor outward: to live documentation, external databases, your GitHub repos, and the real-time web.
The result is an agent that can look up the actual current API signature instead of guessing it, query your production database schema directly, open a PR from the same conversation where you wrote the code, and search the web when it needs information that is not in the codebase.
Setup time
15–30 min for all 6 MCPs
Impact
Fewer hallucinations, less context switching
Cost
All 6 are free or open source
The 6 MCPs every Cursor user should install
These cover the most common gaps in Cursor's default capabilities. Each has minimal setup and immediate, noticeable impact.
Context7
3 min setupModels hallucinate outdated APIs constantly. Context7 eliminates this by grounding every answer in real, current documentation.
npx -y @upstash/context7-mcp
Filesystem
2 min setupMost AI workflows involve reading or modifying files. This MCP is the standard way to give models that access without exposing the full system.
npx -y @modelcontextprotocol/server-filesystem /path/to/allowed/dir
GitHub
5 min setupGitHub is where most code lives. This MCP lets agents interact with that code directly, without copy-pasting between interfaces.
npx @github/github-mcp-server
PostgreSQL
3 min setupDatabase access is one of the most requested AI capabilities. This MCP provides it safely, with schema context that improves query quality.
npx -y @modelcontextprotocol/server-postgres postgresql://user:pass@localhost/db
Brave Search
5 min setupLLMs have a knowledge cutoff. This MCP bridges that gap with live web data directly inside the model context.
npx -y @modelcontextprotocol/server-brave-search
Memory
2 min setupAI models are stateless by default. Memory MCP is the simplest way to give an agent long-term knowledge that carries over between sessions.
npx -y @modelcontextprotocol/server-memory
Start with Context7
How to configure MCPs in Cursor
Cursor reads MCP configuration from a JSON file in your home directory. Here is the step-by-step setup:
Open or create your MCP config
Cursor reads MCP server configuration from this file. If it does not exist, create it.
~/.cursor/mcp.json
Add your first server
Each key in mcpServers is a server name you choose. The command and args fields specify how to start the server process.
{
"mcpServers": {
"context7": {
"command": "npx",
"args": ["-y", "@upstash/context7-mcp"]
}
}
}Restart Cursor and verify
After saving mcp.json, restart Cursor. In Agent mode, you should see the MCP tools listed as available. Test it with a prompt that requires the tool.
Full mcp.json with all 6 MCPs
Copy this into your ~/.cursor/mcp.json to add all six servers at once. Adjust paths and connection strings for your environment.
{
"mcpServers": {
"context7": {
"command": "npx",
"args": ["-y", "@upstash/context7-mcp"]
},
"filesystem": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-filesystem",
"/path/to/your/projects"
]
},
"github": {
"command": "npx",
"args": ["@github/github-mcp-server"],
"env": {
"GITHUB_PERSONAL_ACCESS_TOKEN": "your-token-here"
}
},
"postgres": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-postgres",
"postgresql://user:pass@localhost/mydb"
]
},
"brave-search": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-brave-search"],
"env": {
"BRAVE_API_KEY": "your-brave-api-key"
}
},
"memory": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-memory"]
}
}
}Keep secrets out of your config file
Quick comparison
| MCP | Primary use in Cursor | Setup | API key? |
|---|---|---|---|
| Context7 | Accurate library documentation | 2 min | No |
| Filesystem | Read/write local files | 2 min | No |
| GitHub | Repos, PRs, issues | 5 min | GitHub token |
| PostgreSQL | Database schema & queries | 3 min | Connection string |
| Brave Search | Real-time web search | 3 min | Brave API key |
| Memory | Persist context across sessions | 1 min | No |
Frequently asked questions
Does Cursor support MCP natively?
Yes. Cursor has built-in MCP support. You configure MCP servers in your ~/.cursor/mcp.json file and they become immediately available in Cursor Agent mode.
How do I add an MCP to Cursor?
Open or create ~/.cursor/mcp.json and add your server configuration. Each entry needs a command and optional args and env fields. Restart Cursor after editing the file.
What is the difference between Cursor built-in context and MCPs?
Cursor's built-in context (@ mentions, codebase indexing) is optimized for your current project. MCPs extend Cursor to external services — live documentation, remote databases, GitHub API, web search. They complement each other.
Which MCP should I install first for Cursor?
Context7 is the highest-impact first install. It replaces hallucinated API documentation with real, up-to-date library docs. The improvement in code quality is immediate.
Can MCPs slow down Cursor?
MCPs run as separate processes and only activate when the model calls them. They have no impact on Cursor's editor performance. The only overhead is the latency of the tool call itself, which is typically under 500ms for local MCPs.
See all MCPs for coding
Browse the full list of coding-focused MCPs with rankings, setup guides, and compatibility info.